Health economic evaluation of self-injection of biologics in patients with rheumatoid arthritis using a Japanese real-world web-based survey

Abstract

Background

Biological disease-modifying antirheumatic drugs (bDMARDs) have dramatically improved the quality of life of patients with rheumatoid arthritis (RA); however, concerns regarding their high cost persist. Self-injection (SI) may reduce medical expenses by decreasing the frequency of hospital visits. In this study, we compared the health economic costs of patients with RA who selected SI of bDMARDs and those who did not.

Methods

In this cross-sectional study, we analyzed data from January 2024 using a web-based self-report survey provided by Medilead, Inc. This study included patients with RA who were divided into the SI and non-SI groups. We calculated per visit and total annual out-of-pocket medical expenses for each group.

Results

Among 326 patients with RA, 79 (24.2%; female: 64.6%; mean age: 60.2 years) were treated with bDMARDs. Of these, 65 patients (82.3%) selected SI, and 14 (17.7%) selected non-SI administration. The non-SI group had a higher median frequency of hospital visits than the SI group (12 vs. 6 visits per year). The median out-of-pocket medical expense per visit was higher in the SI group (155.17 USD/visit) than in the non-SI group (86.21 USD/visit). However, the SI group had lower total annual out-of-pocket medical expenses than the non-SI group (948.42 USD/year vs. 1,071.72 USD/year, respectively).

Conclusion

Over 80% of patients with RA selected SI to administer bDMARDs, and their total annual out-of-pocket medical expenses were lower than those of patients who selected non-SI owing to the reduced frequency of hospital visits. The results of our study may provide useful insights into the selection of self-injectable bDMARDs for therapeutic decision-making based on out-of-pocket medical expenses.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-025-12908-1.

Background

Rheumatoid arthritis (RA) is a chronic inflammatory disease. In the last 20 years, the use of biological disease-modifying antirheumatic drugs (bDMARDs) has led to clinical, structural, and functional remission of RA. In Japan, as of January 2025, 12 bDMARDs (infliximab [IFX], infliximab biosimilar [IFX-BS], etanercept [ETN], etanercept biosimilar [ETN-BS], adalimumab [ADA], adalimumab biosimilar [ADA-BS], golimumab [GOL], certolizumab pegol [CZP], ozoralizumab, tocilizumab [TCZ], sarilumab [SAR], and abatacept [ABT]) were approved for RA, with 10 bDMARDs, except for IFX and IFX-BS, approved for self-injection (SI). Although bDMARDs can substantially improve the quality of life of patients, there are concerns regarding their high costs [1, 2].

SI enables patients or caregivers to administer medication at home and has the potential to reduce the following costs during hospital visits: (i) direct healthcare costs, such as medical and drug costs [3]; (ii) direct non-healthcare costs, such as travel costs [4]; and (iii) indirect costs, such as the loss of patients’ or caregivers’ work productivity. For example, from a societal perspective, self-injected subcutaneous depot medroxyprogesterone acetate (DMPA) was found to avert more pregnancies and cost less than health-worker-administered intramuscular DMPA [4]. In patients with inflammatory bowel disease, the annual cost of maintenance treatment with self-injected subcutaneous vedolizumab is 15.0% lower than that of maintenance treatment with health-worker-administered intramuscular vedolizumab [3]. Previous studies evaluating the health economics of RA treatment have focused primarily on the cost-effectiveness of bDMARDs [5–7], with only a few studies exploring the health and economic benefits of SI.

Accordingly, we hypothesized that SI would provide economic health benefits to patients with RA. In this study, we aimed to compare health-related costs between patients with RA who selected SI and those who did not.

Methods

Study design

This cross-sectional study used a web-based self-reported questionnaire for Japanese patients with RA. The participants were patients registered in Medilead, Inc. (Tokyo, Japan) [8–10]. Meadilead Inc. comprises the general Japanese population who agreed to participate in surveys such as questionnaires and interviews. Approximately 4.2 million individuals are registered with Meadilead, Inc. Among all the registrants, eligible participants completed self-reported questionnaires (Supplementary material – Study Questionnaire). This study included 326 participants who were previously diagnosed with RA by a medical doctor. The web-based survey was conducted from January 15, 2024, to January 31, 2024, and the patients responded to the questionnaire only once. We did not collect identifiable information such as names or addresses; all data were anonymized. This study was approved by the Meiji Pharmaceutical University Research Ethics Committee in April 2022 (Approval No. 202144). Informed consent was obtained from respondents using a web form at the beginning of the survey.

Study participants

Participants who responded that RA was their primary disease and were using disease-modifying antirheumatic drugs (DMARDs; lobenzarit, auranofin, salazosulfapyridine, iguratimod, tacrolimus hydrate, leflunomide, bucillamine, methotrexate, actarit, penicillamine, mizoribine, IFX, IFX-BS, ETN, ETN-BS, ADA, ADA-BS, GOL, CZP, TCZ, SAR, ABT, tofacitinib citrate, baricitinib, peficitinib hydrobromide, filgotinib maleate, or upadacitinib hydrate) were identified as the DMARDs group.

In the DMARDs group, we identified users of at least one self-injectable bDMARDs (ETN, ETN-BS, ADA, ADA-BS, GOL, CZP, TCZ, SAR, and ABT) and categorized them into the bDMARDs group.

In the bDMARDs group, patients who selected SI were categorized into the SI group, whereas the remaining patients were assigned to the non-SI group. TCZ and ABT are available as self-injectable subcutaneous and hospital-based intravenous (IV) infusion formulations. In this study, regardless of the formulation used, patients were classified into SI or non-SI groups based on their self-reported method of administration (i.e., whether they selected SI) for subcutaneous (SC) bDMARDs. Accordingly, even if TCZ or ABT was administered intravenously at a medical facility, patients who self-injected any bDMARDs subcutaneously were classified into the SI group. Therefore, IV formulations of TCZ and ABT may have been included in the aggregated bDMARDs data for both the SI and non-SI groups (Tables 1 and 2).

Table 1.

Patients’ characteristics

		bDMARDs group (n = 79)	SI group (n = 65)	non-SI group (n = 14)
Female		51 (64.6%)	42 (64.6%)	9 (64.3%)
Age (Mean ± SD, year)		60.2 ± 12.2	59.3 ± 11.5	67.0 ± 13.5
Disease duration of RA (Mean ± SD, year)		13.1 ± 9.9	13.2 ± 10.3	11.3 ± 8.3
Comorbidities	Hypertension	14 (17.7%)	12 (18.5%)	2 (14.3%)
	Dyslipidemia	11 (13.9%)	8 (12.3%)	3 (21.4%)
	Diabetes mellitus	6 (7.6%)	5 (7.7%)	1 (7.1%)
	Depression	4 (5.1%)	3 (4.6%)	1 (7.1%)
	Respiratory disease	4 (5.1%)	3 (4.6%)	1 (7.1%)
J-HAQ (Mean ± SD)		0.62 ± 0.72	0.61 ± 0.71	0.63 ± 0.75
Employment status	Permanent employment	22 (27.8%)	21 (32.3%)	1 (7.1%)
	Temporary employment	16 (20.3%)	13 (20.0%)	3 (21.4%)
	Unemployed	41 (51.9%)	31 (47.7%)	10 (71.4%)
WPAI (Median [IQR], %)	Absenteeism	0 (0‒0) (n = 38)	0 (0‒0) (n = 34)	0 (0‒0) (n = 4)
	Presenteeism	15 (0‒37.5) (n = 32)	10 (0–40) (n = 29)	20 (0–30) (n = 3)
	Overall work impairment	0 (0‒0) (n = 32)	0 (0‒0) (n = 29)	0 (0‒0) (n = 3)
	Daily activity impairment	20 (0‒40) (n = 79)	20 (10‒45) (n = 65)	20 (0‒32.5) (n = 14)
Annual income (Median [IQR], USD)	Individual	12,068.97 (5,172.41‒24,137.93)	12,068.97 (5,172.41‒24,137.93)	12,068.97 (5,172.41‒24,034.48)
	Household	31,034.48 (18,965.52‒51,724.14)	31,034.48 (18,965.52‒51,724.14)	27,586.21 (15,103.45‒44,827.59)
Coinsurance (Median [IQR], %)		30 (20‒30)	30 (20‒30)	20 (15‒30)
bDMARDs	ETN (including biosimilar)	29 (36.7%)	26 (40.0%)	3 (21.4%)
	TCZ	22 (27.8%)	16 (24.6%)	6 (42.9%)
	ADA (including biosimilar)	12 (15.2%)	10 (15.4%)	2 (14.3%)
	ABT	12 (15.2%)	8 (12.3%)	4 (28.6%)
	GOL	10 (12.7%)	6 (9.2%)	4 (28.6%)
	SAR	7 (8.9%)	4 (6.2%)	3 (21.4%)
	CZP	5 (6.3%)	4 (6.2%)	1 (7.1%)
Medical department	Rheumatology	40 (50.6%)	36 (55.4%)	4 (28.6%)
	Orthopedic surgery	27 (34.2%)	18 (27.7%)	9 (64.3%)
	Internal medicine	13 (16.5%)	11 (16.9%)	2 (14.3%)
	Collagen diseases	11 (13.9%)	9 (13.8%)	2 (14.3%)

Intravenous (IV) formulations of tocilizumab and abatacept may be included in the aggregated bDMARDs data

bDMARDs biological disease-modifying antirheumatic drugs, SI self-injection, SD standard deviation, IQR interquartile range, RA rheumatoid arthritis, J-HAQ Japanese Health Assessment Questionnaire, WPAI work productivity and activity impairment, ETN etanercept, TCZ tocilizumab, ADA adalimumab, ABT abatacept, GOL golimumab, SAR sarilumab, CZP certolizumab pegol

Table 2.

Usage proportion of bDMARDs in each medical department

Medical department	ETN (including biosimilar)	TCZ	ADA (including biosimilar)	ABT	GOL	SAR	CZP
Rheumatology (n = 40)	14 (35%)	9 (22.5%)	7 (17.5%)	7 (17.5%)	5 (12.5%)	4 (10.0%)	3 (7.5%)
Orthopedic surgery (n = 27)	12 (44.4%)	9 (33.3%)	5 (18.5%)	3 (11.1%)	4 (14.8%)	2 (7.4%)	1 (3.7%)
Internal medicine (n = 14)	7 (50.0%)	5 (35.7%)	3 (21.4%)	1 (7.1%)	2 (14.3%)	2 (14.3%)	2 (14.3%)
Collagen diseases (n = 11)	4 (36.4%)	3 (27.3%)	4 (36.4%)	3 (27.3%)	2 (18.2%)	3 (27.3%)	1 (9.1%)

Intravenous (IV) formulations of tocilizumab and abatacept may be included in the aggregated bDMARDs data

Patients who visited multiple medical departments were counted in each department they visited

bDMARDs biological disease-modifying antirheumatic drugs, ETN etanercept, TCZ tocilizumab, ADA adalimumab, ABT abatacept, GOL golimumab, SAR sarilumab, CZP certolizumab pegol

Survey variables

Patient characteristics at the time of survey implementation, including sex, age, duration of RA, comorbidities, Japanese Health Assessment Questionnaire [11], employment status, annual income, coinsurance, work productivity and activity impairment (WPAI) questionnaire, class of bDMARDs used, medical departments visited (rheumatology, orthopedic surgery, internal medicine, and collagen diseases), and the percentage of those who used bDMARDs in each department were collected. Comorbidities were defined as conditions for which hospitalization or outpatient visits occurred within the past year, excluding RA. The WPAI-RA questionnaire comprises six questions as follows [12]. (Q1) Are you currently employed (working for pay)? (Q2) During the past 7 days, how many hours did you miss from work because of problems associated with RA? (Q3) During the past 7 days, how many hours did you miss from work because of any other reason, such as annual leave, holidays, time off to participate in this study? (Q4) During the past 7 days, how many hours did you actually work? (Q5) During the past 7 days, how much did your health problems affect your productivity while you were working? and (Q6) During the past 7 days, how much did your health problems affect your ability to do your regular daily activities other than work at a job? The answers for Q2, Q3, and Q4 are indicated by time (hours), and those for Q5 and Q6 are indicated by scores (0–10). A score of 0 means that health problems had no effect on the patient’s work/daily activities, and a score of 10 means that health problems completely prevented the patient from work/daily activities. We calculated productivity loss as follows: absenteeism (%) = Q2/(Q2 + Q4); presenteeism (%) = Q5/10 × 100; overall work impairment (%) = Q2/(Q2 + Q4) + (1 − Q2/(Q2 + Q4))×(Q5/10)×100; daily activity impairment (%) = Q6/10 × 100. Higher percentages indicate that there was greater work and daily activity impairment and lower productivity.

We collected data on travel time to the hospital, time off for hospital visits, chaperones for hospital visits, out-of-pocket medical expenses per visit, travel costs, and the frequency of hospital visits. We calculated the total annual out-of-pocket medical expenses and limited societal perspective direct costs. The total annual out-of-pocket medical expenses were calculated using the following formula (1 USD = 145 JPY in 2024):

The limited annual societal perspective direct costs were calculated based on the following steps (1 USD = 145 JPY in 2024):

1. For patients whose out-of-pocket payment amounts and co-payments (e.g., 10%, 20%, 30%) were available, we calculated medical expenses per visit from the public healthcare payer’s perspective using the following formula:

For example, if a patient paid 155.17 USD per visit with a 30% co-payment, the total cost was estimated to be 517.23 USD.

• (2)We then calculated the median values for medical expenses per visit, traveling costs, and annual frequency of hospital visits among these patients.
• (3)Finally, we estimated the limited annual societal perspective direct costs using the following formula:

In this study, “limited annual societal perspective direct costs” were estimated by combining direct medical expenses (e.g., drug costs and consultation fees) with direct non-medical expenses, specifically traveling expenses related to hospital visits. This definition is based on internationally accepted frameworks for health economic evaluation, in which adopting a societal or quasi-societal perspective is recommended to include both direct medical and direct non-medical expenses. Although this study does not incorporate indirect costs such as productivity loss or informal care, we aimed to go beyond the payer’s perspective and partially capture the economic burden on patients. Given that SI can reduce the number of hospital visits, the inclusion of traveling costs was considered relevant for assessing the economic benefits of SI.

In addition, we surveyed the level of satisfaction with SI (satisfaction, continuity, implementer, ease of use, reasons for not selecting SI, and the preferred method for future administration). Satisfaction with SI was determined based on five levels: “very satisfied,” “somewhat satisfied,” “neutral,” “somewhat dissatisfied,” and “very dissatisfied.” SI continuity was determined based on the following: “Patient would like to continue SI in the future,” “Patient would like to switch to other methods of administration,” and “Do not know.”

The SI implementer was surveyed based on categories of “Yourself,” “Your family,” “Home health nurse,” and “Other person.”

The ease of SI was determined based on five levels: “very easy,” “somewhat easy,” “neutral,” “somewhat difficult,” and “very difficult.”

Reasons for not selecting SI were determined based on the following survey replies: “patient was unaware of SI”; “patient knew about SI but did not receive information from their doctor”; “patient has used SI in the past, but has now stopped doing so”; “patient has difficulty in SI due to deformed joints”; “patient does not have a good impression of SI”; and “other reasons.” Patients who selected multiple reasons were included.

Outcomes

The primary outcome was the total annual out-of-pocket medical expenses of patients with RA who selected the SI of bDMARDs and those who did not. As a secondary outcome, we compared the limited annual societal perspective direct costs between patients with RA who selected the SI of bDMARDs and those who did not.

Statistical analysis

For the endpoints, the response data are summarized separately for the bDMARDs, SI, and non-SI groups. Summary statistics were calculated for continuous variables, and ratios were calculated for categorical variables. All analyses were performed using R software version 3.2.2.

Results

Patients’ characteristics

This study included 326 patients previously diagnosed with RA by a medical doctor. Among them, 303 (92.9%) and 79 (24.2%) patients were assigned to the DMARDs and bDMARDs groups, respectively (Fig. 1). In the bDMARDs group, 65 patients (82.3%) selected SI (SI group), whereas 14 (17.7%) did not (non-SI group). No patient was undergoing monotherapy with bDMARDs.

Fig. 1.

Patient flowchart. RA, rheumatoid arthritis; DMARDs, disease-modifying antirheumatic drugs; bDMARDs, biological disease-modifying antirheumatic drugs; SI, self-injection

Patient characteristics are presented in Table 1. The mean age of patients in the non-SI group was higher than that of patients in the SI group (67.0 ± 13.5 years vs. 59.3 ± 11.5 years, respectively). Thirty-four patients (52.3%) in the SI group had permanent or temporary employment, which was higher than that in the non-SI group (n = 4; 28.5%).

In the SI group, the median (IQR) scores were as follows: absenteeism, 0% (0–0); presenteeism, 10% (0–40); overall work impairment, 0% (0–0); and daily activity impairment, 20% (10–45).

Conversely, median (IQR) scores for the non-SI group were as follows: absenteeism, 0% (0–0); presenteeism, 20% (0–30); overall work impairment, 0% (0–0); and daily activity impairment, 20% (0–32.5).

Utilization of bDMARDs

The most frequently used bDMARDs were ETN in the SI group (n = 26; 40.0%) and TCZ in the non-SI group (n = 6; 42.9%) (Table 1).

The medical departments with the highest proportion of hospital visits were rheumatology in the SI group (n = 36; 55.4%) and orthopedics in the non-SI group (n = 9; 64.3%) (Table 1).

ETN, TCZ, and ADA were frequently used in all medical departments (Table 2).

Total annual out-of-pocket and limited annual societal perspective direct costs

The SI group had higher median out-of-pocket medical expenses per visit than the non-SI group (155.17 USD/visit vs. 86.21 USD/visit, respectively) (Table 3). However, the SI group had lower total annual out-of-pocket medical expenses than the non-SI group (948.42 USD/year vs. USD 1,071.72 USD/year, respectively) (Fig. 2).

Table 3.

Out-of-pocket and public health care payer’s perspective medical expenses

Perspective of the analysis	Items	bDMARDs group (n = 79)	SI group (n = 65)	non-SI group (n = 14)
-	Travel time to the hospital (Median [IQR], min)	15 (15‒45)	15 (15‒45)	15 (15‒45)
	Time off for hospital visits (Median [IQR], h/visit)	0 (0‒6)	0 (0‒6)	0 (0‒0)
	Chaperone for hospital visits	11 (13.9%)	11 (16.9%)	0 (0.0%)
Out-of-pocket	Medical expenses (Median [IQR], USD/visit) (Direct health care costs)	155.17 (86.21‒241.38)	155.17 (103.45‒241.38)	86.21 (12.93–129.31)
	Traveling costs (Median [IQR], USD/visit) (Direct non-health care costs)	3.03 (0‒6.48)	2.90 (0-6.62)	3.10 (1.93–4.83)
	Frequency of hospital visits (Median [IQR], visit/year) (Number of visits)	6 (4‒12)	6 (4‒12)	12 (10.5‒12)
Public health care payer’s perspective	Medical expenses (Median [IQR], USD/visit) (Direct health care costs)	632.18 (459.77‒1034.48) (n = 69)	775.86 (517.24‒1077.59) (n = 58)	287.36 (258.62‒603.45) (n = 11)
	Traveling costs (Median [IQR], USD/visit) (Direct non-health care costs)	2.90 (0.38‒6.30) (n = 69)	2.83 (0.00‒6.20) (n = 58)	3.45 (2.76‒6.90) (n = 11)
	Frequency of hospital visits (Median [IQR], visit/year) (Number of visits)	6 (4‒12) (n = 69)	6 (4‒12) (n = 58)	12 (6‒12) (n = 11)

bDMARDs biological disease-modifying antirheumatic drugs, SI self-injection, IQR interquartile range

Fig. 2.

Total annual out-of-pocket medical expenses. bDMARDs, biological disease-modifying antirheumatic drugs; SI, self-injection

The median public healthcare payer’s perspective medical expenses per visit were higher in the SI group (775.86 USD/visit) than those in the non-SI group (287.36 USD/visit) (Table 3). The SI group had higher limited annual societal perspective direct costs than the non-SI group (4,672.14 USD/year vs. 3,489.72 USD/year, respectively) (Fig. 3).

Fig. 3.

Limited annual societal perspective direct costs. bDMARDs, biological disease-modifying antirheumatic drugs; SI, self-injection

Regarding the total annual out-of-pocket medical expenses and limited annual societal perspective direct costs, the medical costs of the SI and non-SI groups were reversed.

Fact-finding surveys of SI

Considering satisfaction with SI, a total of 50 patients (76.9%) in the SI group responded “somewhat satisfied” or “very satisfied.” Fifty-two patients (80.0%) responded that they would like to continue SI in the future. Sixty-four patients (98.5%) self-administered the injections.

Regarding the ease of use of SI, 56 patients (86.1%) responded that it was “very easy” or “somewhat easy” (Table 4).

Table 4.

Satisfaction, continuity, implementer, and ease of use of self-injection in the SI group

Survey items	SI group (n = 65)
Level of satisfaction with SI
Very satisfied	15 (23.1%)
Somewhat satisfied	35 (53.8%)
Neutral	10 (15.4%)
Somewhat dissatisfied	5 (7.7%)
Very dissatisfied	0 (0.0%)
Continuity of SI
Patient would like to continue SI in the future	52 (80.0%)
Patient would like to switch to another method of administration	4 (6.2%)
Do not know	9 (13.8%)
SI implementer
Yourself	64 (98.5%)
Other person	1 (1.5%)
Ease of SI use
Very easy	29 (44.6%)
Somewhat easy	27 (41.5%)
Neutral	8 (12.3%)
Somewhat difficult	1 (1.5%)
Very difficult	0 (0.0%)

SI self-injection

In the non-SI group, 10 patients (71.4%) had a negative impression of SI. Among these, six patients (42.9%) did not select SI owing to “not having a good impression of SI”; of the remaining four (28.6%), one patient each selected “patients was unable to administer the injection by himself/herself due to a lack of strength, and patient was afraid to inject himself/herself,” “patient worried about forgetting to inject,” “patient disliked SI by himself/herself” and “patient was feeling anxious about SI”. In addition, one patient each selected “patient did not select SI according to the doctor’s instruction” and “patient did not need to select SI”. Furthermore, one patient each selected “patient has used SI in the past but has now stopped from doing so” and “patient has difficulty with SI due to deformed joints”. Three patients (21.4%) knew about SI but had not received information from their doctors (Table 5). Thirteen patients (92.9%) preferred to continue with their current method of administration and did not want to switch to SI.

Table 5.

Reasons for not selecting self-injection in the non-SI group

Reasons for not selecting SI	Non-SI group (n = 14)
Patient does not have a good impression of SI	6 (42.9%)
Other reasons	6 (42.9%)
Patient was unable to administer the injection by himself/herself due to a lack of strength, and the patient was afraid to administer the injection himself/herself	1 (7.1%)
Patient worried about forgetting to inject	1 (7.1%)
Patient disliked SI by himself/herself	1 (7.1%)
Patient was feeling anxious about SI	1 (7.1%)
Patient did not select SI according to the doctor’s instruction	1 (7.1%)
Patient did not need to select SI	1 (7.1%)
Patient knew about SI but did not receive information from their doctor	3 (21.4%)
Patient has used SI in the past but has now stopped from doing so	1 (7.1%)
Patient has difficulty with SI due to deformed joints	1 (7.1%)

SI self-injection

Discussion

In this study, we found that over 80% of patients with RA using bDMARDs selected SI. Although the SI group had higher out-of-pocket medical expenses per visit than the non-SI group, total annual out-of-pocket medical expenses decreased. While SI reduces the frequency of hospital visits, patients may be prescribed bDMARDs for several months in a single visit in anticipation of potential natural disasters or limited access to healthcare services. Healthcare providers must work with the government to devise easy-to-understand measures to inform residents of the relevance of stockpiling regular medications for disasters and conduct educational activities for chronic diseases [13]. Accordingly, the medication cost per visit may increase, contributing to higher direct medical costs and, consequently, greater overall societal costs. However, Japan’s high-cost medical expense benefit system sets an upper limit on out-of-pocket payments based on individual income levels. This system may have contributed to keeping out-of-pocket payments relatively low in the SI group despite higher medical costs, potentially explaining the observed discrepancy. Based on these results, the medical expense evaluation of SI for patients with RA using biologics may vary depending on the analytical perspective. This study was limited to results based on a web-based patient survey and did not provide an appropriate health economic evaluation from the perspective of public healthcare payers. Therefore, future studies should consider large claims data to evaluate medical expenses associated with SI more accurately from the public healthcare payer’s perspective.

The difference in the number of hospital visits between the SI and non-SI groups may be explained by differences in employment status according to age. In this study, the mean age of participants in the non-SI group was higher than that of participants in the SI group. Older outpatients reportedly prefer bDMARDs administration by medical staff at medical facilities, whereas younger patients prefer SI [14]. This may be because a lower proportion of older patients was employed, making it easier for them to secure time for hospital visits. In this study, the proportion of patients with permanent or temporary employment was lower in the non-SI group than in the SI group. Therefore, the non-SI group had more frequent total annual hospital visits than the SI group.

Informal care is unpaid care provided by families or non-family volunteers to fulfill patients’ needs to accomplish activities of daily living (ADLs), such as bathing, dressing, and eating, or instrumental ADLs, such as shopping, cooking, and money management [15]. From a societal perspective, it is important to consider health economics, including informal care costs. A study conducted in the United Kingdom reported that caregivers of patients with moderate RA required time off work and reduced working hours to provide care [16]. In the present study, 11 patients (16.9%) in the SI group required chaperones for hospital visits (Table 3). We observed that SI reduced the frequency of hospital visits, which may have reduced the frequency of chaperone hospital visits. Additionally, the median travel time to the hospital per visit was only 15 min for the SI and non-SI groups. However, SI reduced economic losses by decreasing the frequency of hospital visits, resulting in reduced travel time. To evaluate the health and economic benefits of SI from a societal perspective, it is crucial to consider the economic losses incurred by caregivers and patient travel times to hospitals.

The results of our study indicate that patients were satisfied with their current method of administration, which is consistent with the findings of a previous study [17]. Conversely, patients or their caregivers may experience anxiety or fear regarding SI [18–21]. Negative past experiences with SI can affect future SI experiences [18]. In our study, more than 70% of patients who did not select SI had a negative impression of it, and we observed a similar trend to that of Smith et al. This study showed that over 20% of patients who had not selected SI did not receive information from their doctors or were unaware of SI. These results indicate that providing appropriate information may encourage more patients to switch to SI.

Herein, we detected differences in the distribution of medical departments between the SI and non-SI groups. Specifically, most patients in the SI group visited rheumatology departments (36 patients; 55.4%), whereas those in the non-SI group tended to visit orthopedic departments more frequently (9 patients; 64.3%). These differences may have influenced both the decision to use SI and the subsequent economic outcomes. Rheumatologists are generally well-experienced in bDMARDs therapy and are more likely to provide active guidance and information regarding SI. Conversely, orthopedic departments, even when rheumatology specialists are involved, typically focus on structural and procedural treatments of joints, and may provide less information or support related to self-injection. Such differences across medical departments may affect patients’ choices regarding the method of bDMARDs administration, ultimately leading to variations in visit frequency and healthcare costs.

However, this study has some limitations that must be addressed. First, we used data from a web-based, self-reported survey, and the possibility of RA misclassification cannot be disregarded. Second, this study was limited to patients who could participate in a web-based survey and had relatively high information technology literacy and interest in their disease. Third, we could not assess clinical data on disease and physical activity. Fourth, because this study was based on only one survey, causal relationships could not be investigated. Fifth, this study was conducted without distinguishing between SI and intravenous formulations; therefore, for patients who selected TCZ and ABT as bDMARDs, the use of SI and intravenous formulations may have been included. Sixth, the term “negative impression” reflects a psychological reluctance or fear of SI; however, it remains unclear whether these impressions are associated with prior adverse experiences. Additionally, owing to the structure of the questionnaire, data on the types of bDMARDs previously used and whether they were administered via SI or non-SI methods were not collected. This limitation may have affected the interpretation of the patient’s attitudes toward SI. Finally, this study has the potential for recall bias owing to its reliance on self-reported data. Information such as the frequency of hospital visits over the past 12 months was collected using a web-based questionnaire, which may have led to variability in the accuracy of responses. However, the questionnaire employed a multiple-choice format rather than open-ended responses to minimize the respondent burden and reduce the impact of recall bias. Furthermore, patients with RA experiencing stable disease activity tend to attend regular follow-up visits once a month or every two months, which may lead to accurate recall.

Conclusion

Over 80% of patients with RA using bDMARDs selected SI, which lowered their total annual out-of-pocket medical expenses by reducing the frequency of hospital visits. The results of our study may provide useful insights into the selection of self-injectable bDMARDs for therapeutic decision-making based on out-of-pocket medical expenses.

Abbreviations

ABT

Abatacept

ADA

Adalimumab

ADA-BS

Adalimumab biosimilar

ADLs

Activities of daily living

bDMARDs

Biological disease-modifying antirheumatic drugs

CZP

Certolizumab pegol

DMARDs

Disease-modifying antirheumatic drugs

DMPA

Depot medroxyprogesterone acetate

ETN

Etanercept

ETN-BS

Etanercept biosimilar

GOL

Golimumab

IFX

Infliximab

IFX-BS

Infliximab biosimilar

IQR

Interquartile range

IV

Intravenous

J-HAQ

Japanese Health Assessment Questionnaire

RA

Rheumatoid arthritis

SAR

Sarilumab

SC

Subcutaneous

SI

Self-injection

TCZ

Tocilizumab

WPAI

Work productivity and activity impairment

Authors’ contributions

K.T., E.T., R.S. and M.A. conceived the study. All authors contributed to study design and interpretation. Data analysis was performed by K.T. and Y.M. The first draft of the manuscript was written by K.T., and all authors commented on previous versions of the manuscript. All the authors have read and approved the final version of this manuscript.

Funding

The authors received no funding for this study.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study was approved by the Meiji Pharmaceutical University Research Ethics Committee in April 2022 (Approval No. 202144). The study complied with principles for ethical research in the Declaration of Helsinki, and informed consent was obtained from all the participants using a web form at the beginning of the survey.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.